Development of an Intelligent Software Platform for Forecasting Market Trends and Stock Prices Based on the Analysis of Big Financial Data and News Feeds

Abstract

Currently, there is a fundamental problem of information noise and high data dimensionality in financial markets. Traders and automated systems face the problem of the non-stationary nature of stock prices, which makes traditional methods of market analysis and forecasting ineffective. This study presents the development of an intelligent software platform for short-horizon stock price movement forecasting based on financial indicators and news sentiment analysis. The system integrates feature engineering, dimensionality reduction via Principal Component Analysis (PCA), and machine learning classification models. A multi-horizon evaluation (1-day, 3day, and 5-day ahead prediction) was conducted to assess time-scale sensitivity. Experimental results indicate that predictive performance remains close to random classification levels (ROC-AUC$\approx 0.47 - 0.51$) across all horizons. These findings highlight the inherent stochastic nature of short-term financial markets and confirm the difficulty of extracting stable directional signals from publicly available information.

More by Nurzhamal Kashkimbayeva

• Evidence-Linked Measurement of Indirect Prompt Injection in Web Security Agents: Paired Trials and Application-Specific Evidence AuditsTechnologies · 2026
• Kazakhstan Agri Recommender — Data Collection PipelineZenodo · 2026
• Replication Data and Code for “Added Predictive Value and Temporal Transferability of MODIS and ERA5-Land Data for District-Level Wheat Yield Forecasting”Harvard Dataverse · 2026
• Multimodal Data Fusion for District-Scale Crop Yield Prediction in Northern KazakhstanIEEE International Conference on Smart Information Systems and Technologies · 2026